Abstract
Abstract. This research incorporates the correlated k distribution BCC-RAD radiation model into the climate model BCC_AGCM2.0.1 and examines the change in climate simulation by implementation of the new radiation algorithm. It is shown that both clear-sky radiation fluxes and cloud radiative forcings (CRFs) are improved. The modeled atmospheric temperature and specific humidity are also improved due to changes in radiative heating rates, which most likely stem from the revised treatment of gaseous absorption. Subgrid cloud variability, including vertical overlap of fractional clouds and horizontal inhomogeneity in cloud condensate, is addressed by using the Monte Carlo Independent Column Approximation (McICA) method. In McICA, a cloud-type-dependent function for cloud fraction decorrelation length, which gives zonal mean results very close to the observations of CloudSat/CALIPSO, is developed. Compared to utilizing a globally constant decorrelation length, the maximum changes in seasonal CRFs by the new scheme can be as large as 10 and 20 W m−2 for longwave (LW) and shortwave (SW) CRFs, respectively, mostly located in the tropics. The inclusion of an observation-based horizontal inhomogeneity of cloud condensate has also a significant impact on CRFs, with global means of ~ 1.5 W m−2 and ~ 3.7 Wm−2 for LW and SW CRFs at the top of atmosphere (TOA), respectively. Generally, incorporating McICA and horizontal inhomogeneity of cloud condensate in the BCC-RAD model reduces global mean TOA and surface SW and LW flux biases in BCC_AGCM2.0.1. These results demonstrate the feasibility of the new model configuration to be used in BCC_AGCM2.0.1 for climate simulations, and also indicate that more detailed real-world information on cloud structures should be obtained to constrain cloud settings in McICA in the future.
Highlights
Radiation process is crucial for climate simulations
This is due to improvements in both the revised cloud optics and the net clear-sky fluxes calculated by the new radiation scheme
The results show that the new scheme markedly improves the representation of the SW and LW radiation budget for both clear-sky and all-sky conditions, whether in the global mean or in geographic distribution
Summary
Radiation process is crucial for climate simulations. Over the past 2 decades, a lot of progress has been made in atmospheric radiation. SW and longwave (LW) CRFs showed significant changes compared to the traditional PPH and maximum-random overlap setup, but the magnitude of changes depends on the cloud scheme utilized All of these studies have emphasized the importance of faithfully addressing subgrid cloud variability in GCMs. To make the representation of subgrid cloud properties flexible and modularized and to maintain computational efficiency, a scheme named the Monte Carlo Independent Column Approximation (McICA) method was developed (Pincus et al, 2003). Most attention will be paid to the cloud-type-related decorrelation length by comparing with the results of using a globally constant value This preliminary work aims to document the impact of the modifications in cloud-radiation process on simulated climate and the model response to these changes and thereby provide suggestions for future development.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
